Arno Solin (Aalto University, Finland)
Low-cost and noisy sensor sources in modern smartphones introduce both interesting possibilities for new applications, and challenges for inference methods needing to cope with the low-quality data. This talks presents applications (and computational challenges) inspired by the sensors available in a standard iPhone: indoor positioning based on the compass sensor (magnetometer), inertial navigation driven by the IMU (accelerometer and gyroscope), and visual-inertial navigation utilizing the phone camera and IMU. The underlying methods are based on stochastic differential equation (SDE) models and Gaussian process based machine learning methods.
see also the list of SFB seminar events
